Converting to BLOB in AL32UTF8 destroys German Umlaut characters

Question

Test Case.
Two database
(A) NLS_CHARACTERSET= WE8MSWIN1252
(B) AL32UTF8
I have used the following function to convert CLOB data to BLOB
create or replace function clob_to_blob (p_clob_in in clob)
return blob
is
v_blob blob;
v_offset integer;
v_buffer_varchar varchar2(32000);
v_buffer_raw raw(32000);
v_buffer_size binary_integer := 32000;
begin
  if p_clob_in is null then
    return null;
  end if;
  DBMS_LOB.CREATETEMPORARY(v_blob, TRUE);
  v_offset := 1;
  FOR i IN 1..CEIL(DBMS_LOB.GETLENGTH(p_clob_in) / v_buffer_size)
  loop
    dbms_lob.read(p_clob_in, v_buffer_size, v_offset, v_buffer_varchar);
    v_buffer_raw := utl_raw.cast_to_raw(v_buffer_varchar);
    dbms_lob.writeappend(v_blob, utl_raw.length(v_buffer_raw), v_buffer_raw);
    v_offset := v_offset + v_buffer_size;
  end loop;
  return v_blob;
end clob_to_blob;If i input ÄÖÜ to the function,In WE8MSWIN1252 the returning BLOB looks ok, but in AL32UTF8 the BLOB's characters comes out like ÄÖÜ.
Now if i were to save the values (CLOB and BLOB) in both occations the CLOB will be stored without problem, but BLOB will be saved incorrectly only in AL32UTF8 environment.
The only difference between the two db's are the characterset.
Is this a know limitation or am i doing something wrong.
PS. I have seen similar behaviour when using dbms_lob.converttoblob
thank you

Answer 1

When describing the problem you missed one very important point: how do you look at the content of the BLOB to tell if it is OK or not? Remember that observation usually influences the experiment results ;-)
The sequence like ÄÖÜ looks like AL32UTF8 encoding of umlauts viewed with a WE8MSWIN1252 viewer. If you look with a viewer that expects WE8MSWIN1252 at a BLOB that contains AL32UTF8 text (which is the case with an AL32UTF8 database), then this is what you should expect.
-- Sergiusz